President Trump has rallies. In 2016, he closed out the general election with a legitimately impressive string of rallies throughout the swing states. He won the presidency, though not the popular vote, a development that seemed to cement his sense that he knew what was needed to win. So when 2018 rolled around, Trump again closed out the midterms with a flurry of rallies … only to see his party get walloped. It’s not even that Democrats won back the House and took a majority of the Senate seats that were up for grabs. It’s also that Trump’s rallies made little discernible difference in polling and may even have cost his party seats by energizing the opposition.
The thing about rallies, though, is that Trump obviously loves them. He gets to surround himself with his most fervent supporters and have them cheer for him for an hour, like a stand-up comedian performing at a nightclub filled only with his best friends and wildly intoxicated joke enthusiasts. Trump can credit the rallies with delivering 2016 for him — and use that as an argument for holding as many rallies as possible.
For the past week, Trump hasn’t been able to hold any rallies, though, given that he was diagnosed with the coronavirus-caused disease that he keeps insisting will soon simply disappear. The frustration he’s felt at being confined to the hospital and the White House is palpable, a cosmic cloud darkening his administration. But after getting a thumbs up from his personal physician, Trump will head back out on the stump this week.
So far, his rallies haven’t done much to move the race. Since his rally in Tulsa in late June — the first after he gave himself a political thumbs-up to resume rallies despite the ongoing pandemic — his poll numbers in states where he has held rallies haven’t moved much. On average, his poll numbers in a state haven’t moved from two weeks before the rally to two weeks after. His poll numbers in any state relative to national polling (a metric used to eliminate the drag on his state numbers from national changes) have improved 0.1 percentage points on average. Which is essentially no change at all.
There are times when his poll numbers have surged after a rally. In the two weeks after his rally in Pennsylvania on Aug. 20, for example, his poll numbers in the state shot up two points. But that was probably a function of the Republican convention that began less than a week after his rally.
On the plus side, there’s at least more of a reason to hold a rally at this point than there was in June or July. A central reason to hold such an event is to generate media attention and, more concretely, to push people to vote. In July, there’s no use in urging people to vote, since they can’t. Now, though, voters in most states can cast early ballots by absentee or in person. And while Trump is heavily invested in undermining confidence in early voting — reserving the right to falsely claim that it’s all corrupt, should he be in a tenuous position on Election Day — he hasn’t really made encouraging voters to cast ballots a central part of his recent rallies.
In Duluth, Minn., for example, Trump gave a shorter-than-usual speech at the end of last month, perhaps because he was already starting to feel the effects of the coronavirus. The only enticement to vote was a generic one — “go out and vote” — without the expected reinforcement of making a plan to vote or reminders of different options to do so. Trump doesn’t normally work standard political patter into his speeches, and it might be the case that attendees received such instructions from other speakers or in other ways. But the end result was a get-out-the-vote rally in which Trump simply said “get out and vote” three times, during an event held on the border with a state, Wisconsin, where early voting hadn’t even started.
Since that point, Trump’s standing in the polls has eroded. This is probably a function both of his aggressive approach to the first presidential debate and to the fact that, given his illness, the race has been laser-focused on the pandemic, which isn’t to his benefit. Recent polling, including from The Washington Post and our partners at ABC News, has shown that Trump’s seen a broad erosion of support from women and older voters.
Trump’s campaign has an answer to that, as articulated in a call with reporters Monday. Sure, their support with older voters is weaker than it was in 2016, but that would be “offset by gains in certain voting populations,” such as Black and Hispanic voters.
If we compare our most recent poll to 2016 exit polls compiled by Edison Media Research, you can see the problem with that strategy. Among non-White voters, Hillary Clinton held a 74 percent to 21 percent advantage in 2016. That group made up 29 percent of all voters, according to Edison, meaning that Clinton got 21.5 percent of all votes cast from non-White voters (74 percent of 29 percent) and Trump got 6.1 percent. That’s a 15-point advantage for Clinton.
In our new poll, former vice president Joe Biden does about as well as Clinton did four years ago. If we assume the same turnout from non-White voters, Biden doesn’t actually see a decline in support relative to Clinton. Applying the same math to older voters, we see that Biden gains 1 percent of the electorate relative to Clinton.
65 and older
Percent of voters
Clinton share of voters
Biden share of voters
That’s non-White voters overall. Data from Pew Research, including a validated look at the 2016 electorate and its most recent national poll, released last week, shows a similar net result. Here, Trump gains 0.4 percent of the electorate from Black and Hispanic voters — but loses 2.4 percent from older voters.
65 or older
Percent of voters
Clinton share of voters
Biden share of voters
There’s another problem for Trump here. The numbers above assume a static distribution of voters. If the percentage of voters who are Black jumps from 10 percent to 12 percent, for example, Biden’s share of the electorate from Black voters jumps to 9.9 percent, an increase over 2016 of 1.4 points.
Of course, there’s also overlap between these groups, making a direct comparison problematic. (There are lots of Black voters over 65, for example.) The numbers do suggest, though, that the trend is less of a wash than Trump’s campaign would like us to think.
Particularly if you look at the shifts among women. Using similar math as the calculations above to compare the Post-ABC poll to 2016 exit polls, Biden picks up 5 percent of the electorate. (Pew’s data, showing a closer race among women, is much less pronounced.)
The overarching question is where Trump picks up any ground at all. As we noted last week, the number of voters who haven’t made up their minds is well below the figures for 2016. Voters skeptical of both Trump and Biden indicate that, unlike four years ago, they’re going to break against the president. His rallies aren’t reshaping the polling in swing states to any noticeable degree. It’s just stasis, which is the ongoing story of the 2020 election.
2016 taught everyone the value of saying, “Of course, things could change.” So: Of course, things could change. If the Trump campaign knows how to make that happen, it’s not immediately obvious.
The good news for Trump, at least, is that he gets to hold rallies anyway.